DP
RIETI Discussion Paper Series 10-E-012
Firm Size and Business Startup Reasons of Japanese Workers
TSUCHIYA Ryuichiro
Ritsumeikan Asia Pacific University
The Research Institute of Economy, Trade and Industry
http://www.rieti.go.jp/en/
RIETI Discussion Paper Series 10-E-012
February 2010
Firm size and business startup reasons of Japanese workers1
Ryuichiro TSUCHIYA2
Ritsumeikan Asia Pacific University
Abstract
Small firms are more likely to produce entrepreneurs than large firms.
This study empirically examines a potential mechanism that might
explain this phenomenon, observed in previous research, using Japanese
survey data of employees planning to start businesses. The data contain
information on employer, job, and personal characteristics and also
indicates the reasons for starting the businesses. The results from a
principal component analysis of various startup reasons identify four
separate component factors that account for 70 percent of variance: a
need for self-fulfillment, a need for flexibility in work schedule, a need to
solve a career problem, and a need to secure a livelihood. I empirically
examine the relationship between rating scores for these factors and the
size of employers. The results from multivariate regression models
indicate that the score for a need to solve a career problem was
significantly higher for those working for small firms, while none of the
other three factors are significantly different between employees of large
and small firms. In addition, the results also suggest that the relationship
between the need to solve a career problem and employment of small
firms is associated with the tendency for middle managers. The
implications of these findings for researchers and policy-makers are
discussed.
Key words: Entrepreneurship; Startup reasons; Firm size
JEL classification: M13; L26
RIETI Discussion Papers Series aims at widely disseminating research results in the form of
professional papers, thereby stimulating lively discussion. The views expressed in the papers are
solely those of the author(s), and do not present those of the Research Institute of Economy,
Trade and Industry.
1
This research is a part of the project on "Study of Entrepreneurs and Latent Entrepreneurs" at the
Research Institute of Economy, Trade and Industry (RIETI). I am grateful to RIETI for generous
financial support. I am also grateful to Takehiko Yasuda and the participants of the Discussion Paper
Workshop held at RIETI for their helpful comments.
2
College of International Management, Ritsumeikan Asia Pacific University, 1-1, Jumonjibaru,
Beppu, Oita 8748577, Japan. Email: rtsuchi@apu.ac.jp
Business startup reasons
Firm size and business startup reasons of Japanese workers
Ryuichiro Tsuchiya
There has been general agreement about the importance of new firm creation and its desirable
effects on economies. Since Birch’s (1979) study on job creation by small businesses, a
considerable amount of research has confirmed his findings that small businesses are a major
source of job creation. However, Acs, Armington, and Robb (1999) found that there is a net
destruction of existing jobs among older businesses whether small, medium, or large. This
suggests that new businesses, not small businesses themselves, provide a major force for
creating new jobs.
Given the economic significance of new firm creation, exploration of the factors that
affect the supply and demand of entrepreneurs has been of practical, as well as academic,
interest. Some research has sought to identify the environment in which new firm formation
is more likely to occur. Key environmental factors that have been argued to lead to the
formation of new firms include technological change (Tushman and Anderson, 1986),
industry dynamics (Klepper, 2002; Sonobe, Kawakami, and Otsuka, 2003), wage differentials
between self and paid employment (Rees and Shah, 1986), market structure (Acs and
Audretsch, 1990), and regional infrastructure (Fritsch, 1992; Okamuro, 2006). On the other
hand, the other school of thought has focused on individual-level factors, including
propensity to take risk, tolerance for ambiguity, need for achievement, and decision-making
patterns (Begley and Boyd, 1987; Busenitz and Barney, 1997) which differentiate firm
founders from the rest of society. The school has also focused on individual differences in
non-psychological characteristics (Bates, 1990; Harada, 2003; Honjo, 2004) and differences
in founding and growth strategies (Yasuda, 2005; Newbert, Kirchhoff and Walsh, 2007).
In these streams of research, recent years have seen a growing interest on the linkage
between the properties of firms and the likelihood of employees leaving to start their own
2
Business startup reasons
businesses. To date, empirical research with regional data has found empirically strong and
robust relationships between the share of small firms in the total employment of region and
the formation rate of new firms (Fritsch, 1992; Reynolds et al., 1994; Gerlach and Wagner,
1994; Spilling, 1996). However, the underlying mechanism for this observation remained
unclear. Only relatively few researchers have explored the effect of firm size on transition
rate at the level of individual (Gompers, et al., 2005; Dobrev and Barnett, 2005; Sørensen,
2007). Despite a growing number of studies, research to date has offered little explanation as
to why small firms are more likely to produce entrepreneurs.
There are important reasons for why it is desirable to give explanation to the
phenomenon. First, exploring a mechanism of the phenomenon may carry policy implications
for policy-makers seeking to promote entrepreneurship among workers. It is natural to expect
that small firms produce more entrepreneurs than do large firms, given that small firms
account for 66 percent of total employment in Japan (Small and Medium Enterprise Agency,
2009). What added to this direct observation is that Yahata’s (1994) study using the database
of the Employment Status Survey, though largely descriptive, revealed that the category of
firms with less than 30 workers were producing a disproportionate number of newly
self-employed persons in Japan. Thus, gaining a better understanding of why small firms are
good generators of entrepreneurs could also contributes to the debate on the role of small
business in Japanese economy. Second, employers could benefit from learning about why
some employees leave to start their own businesses. Employers who seek to retain crucial
employees could be better informed to identify potential quitters who wish to setting up their
own businesses which will likely be direct competitors to themselves.
In an attempt to explore the underlying mechanism effectively, I adopted a
questionnaire survey designed specifically for employees who plan to start their businesses,
which enabled me greatly to identify, among others, their several motives for starting
3
Business startup reasons
businesses, and the effect on the size of employers on these motives. A particular
characteristic of the survey design adopted in this paper, asking employees planning to start
businesses about reasons while they are still in the pre-startup phase, helped overcome a
methodological barrier that researchers often face when attempting to identify a major startup
motive. A serious concern has been the problem of retrospection, where researchers interview
entrepreneurs about startup reasons long after they got into business. Such retrospective
inquiry has been criticized that an accurate measurement is often not possible due to loss of
memory and a bias that makes people justify their behaviors. There is also the possibility of
survival bias, as some may have dropped out of track for business ownership as time passes.
To date, only few studies have successfully solved this problem (Carter, Gartner, Shaver, and
Gatewood, 2003; Schjoedt and Shaver, 2007). For the research on entrepreneurship in Japan,
relatively fewer studies have explored startup reasons. Some of these include Small and
Medium Enterprise Agency (2006) but the study relied on retrospective questions. This
research is capable of minimizing this problem, as I examined the startup reasons while
respondents were still in the initial stage of setting up businesses.
THEORETICAL PERSPECTIVE
The section below reviews related literature, and introduces four theoretical perspectives that
might explain the relationship between the size of firms and the tendency of employees to
start their businesses. Testable hypotheses are proposed for each perspective.
Review of the Literature
Parker (2009) examined relative effectiveness of three alternative theories that could help
explain why small firms are more likely to produce entrepreneurs: the transmission theory;
blocked mobility theory; and self-selection theory. The transmission theory is based on the
4
Business startup reasons
notion that it is easier for small firms than large firms to create opportunities to learn critical
skills, attitudes, knowledge, resources, and social connections that help develop a career in
entrepreneurship. On the other hand, the self-selection theory is based on the observation that
persons with particular personality are more likely to choose to become employees of small
firms, and the same persons are also more likely to start their own businesses sometime in the
course of their careers. The third theory, blocked mobility theory assumes the presence of
“dual labor market”, arguing that a primary motive of employees of small firms to start their
businesses is frustrations with their career opportunities.
Using a longitudinal database of the British labor market, Parker (2009) argued that
neither the transmission theory nor the blocked mobility theory received compelling
empirical support, whereas there was an indication of self-selection behavior. On the other
hand, Sørensen (2007) tested empirically the self selection theory using a longitudinal
database of the Danish labor market. The results from career tracking models provided
evidences to suggest that the effect of firm size is not the spurious consequence of
self-selection behavior. Finally, Tsuchiya (2009) examined empirically whether employees of
small firms are pulled toward opportunities, or pushed by frustration and dissatisfaction with
a job, by using a repeated cross-sectional database of Taiwan’s labor market. The results of
the study suggested that a dominant motive of those working for small firms is pull-based
rather than push-based.
Interpretation of some inconsistency among findings of these studies is complicated
in part by the fact that it is very difficult for researchers to distinguish firm founders into
several distinct categories according to a major motive for starting businesses. For example, a
person who plans to pursue a chance for self fulfillment aggressively might also be frustrated
with a current job to the similar extent. The coexistence of competing motives within the
same person has been criticized on methodological grounds. This issue is reflected in the
5
Business startup reasons
research design adopted by this paper. In this paper, I first subjected the data of subjective
ratings that respondents gave for various items of startup reasons to a principal component
analysis to identify major motive factors of startup, and then tested the effects of firm size on
identified motives by using a multivariate regression model, which are able to minimize the
problem of coexistence of multiple motives.
Yet, another issue that researchers should consider is institution of labor markets and
labor relations. Studies using regional data have suggested that the intensity of new firm
formation in a country is dependent on the degree of legal protection of job security and the
mobility of labor (Choi and Phan, 2006; Van Stel, Storey, and Thurik, 2007). Japanese labor
market is often described as very rigid and inflexible, in which employment protection of
regular employees and preferences of recruiting of large firms for new graduates have
suppressed the expansion of external labor markets. These labor institutions create barriers
for the entry of employees of small firms into internal labor markets of large firms. It is likely
that startup motives of employees in Japanese small firms are somehow different from those
of employees of small firms located in more dynamic and flexible national labor markets.
Thus, cross-country comparison of studies should be interpreted with caution.
Transmission Theory
The essential argument of the transmission theory is that small firms create learning
opportunities for potential entrepreneurs in at least three important ways. First, working in
small organizations requires flexibility of approach (Dobrev and Barnett, 2005; Sørensen,
2009). This may influence the attitudes and mind-set of members of organizations in ways
that make them more likely to pursue uncertain opportunities in entrepreneurship. Second,
experience of working for small firms may facilitate the development of critical skills
necessary for running new organizations, and may therefore raise the expected value of career
6
Business startup reasons
opportunities in entrepreneurship. If successful entrepreneurship require a command of a
wide variety of roles and responsibilities, then persons with diverse experience and expertise
will find entrepreneurial opportunities more attractive (Lazear, 2005). The diversity of work
experiences should be on average higher among employees of small firms, implying that the
rates of entry into business ownership will be higher among them (Sørensen, 2009). Finally,
small firms may give employees more frequent and emotional contacts to a network of
customers and suppliers, which may increase the chance to identify market opportunities and
capacity to mobilize social resources (Dobrev and Barnett, 2005; Sørensen, 2009). These
arguments suggest that employees of small firms are more likely to emphasize the aspect of
personal fulfillment as compared to those of large firms. Thus, I propose the following:
Hypothesis 1a: For employees who wish to start new businesses, those
working for small firms are more likely than those working for large firms
to stress a need for personal fulfillment as a major startup motive.
Blocked Mobility Theory
Another explanation is based on the blocked mobility theory, arguing that marginal workers
excluded from attractive job opportunities available in the internal labor markets of large
firms find a career in entrepreneurship as a solution to the problem. The organizational
structure of small firms is flat relative to that of large firms, which makes career advancement
within organization less attractive as employees climb career ladder (Dobrev and Barnett,
2005). Thus, long-tenured employees of small firms often try to move to other firms but face
obstacles in moving to larger firms. In Japan, these obstacles are more problematic because of
the segmentation of labor markets and the preferences of large firm recruiting for new
graduates. These arguments lead to the following hypothesis:
7
Business startup reasons
Hypothesis 1b: For employees who wish to start new businesses, those
working for small firms are more likely than those working for large firms
to stress a lack of esteem from the employer as a major startup motive.
Self-selection Theory
The self-selection theory argues that people choose jobs consistent with their values to work,
and an innate basis for autonomy that will be likely to lead people to start own businesses
also has an effect on a preference for seeking jobs at small firms (Sørensen, 2007). These
arguments lead to the following hypothesis.
Hypothesis 1c: For employees who wish to start new businesses, those
working for small firms are more likely than those working for large firms
to stress a need for autonomy in work as a major startup motive.
Necessity Theory
What could be added to this set of theories is that the reason that employees of small firms
start businesses is necessity of earning a livelihood. On average small firms pay more
variable wages than large firms, in part because the latter are more diversified than the former
(Parker, 2009). This makes small firms less able to reduce the risk of business cycle shock,
and therefore employees face the volatility of income. Thus, employees of small firms are
more vulnerable to the risk of poverty. This leads to the argument that the high rate of
transition observed for employees of small firms is simply reflecting a high rate of failure for
employees of small firms to earn a living.
8
Business startup reasons
Hypothesis 1d: For employees who wish to start new businesses, those
working for small firms are more likely than those working for large firms
to stress necessity to earn a livelihood as a major startup motive.
RESEARCH METHODOLOGY
Data
Testing hypotheses proposed in the previous section will be challenging, requiring data on
individuals who are in the initial stages of starting businesses. Screening from the population
at large to identify people who might be starting businesses needs considering a balance
between costs and benefits, because transition to business ownership is generally a rare event3.
For this reason, I chose not to attempt to collect a representative sample, and instead focused
on screening from a more refined set of population who would be likely to experience
entrepreneurial transition. Although this approach obviously limits the generalizability of
possible results of this study, it improves the ability of this study to achieve depth of data on
startup reasons.
The database for this study was on-line survey responses collected from 28,495
participants of Yahoo! Research who live in Japan. Candidates of nascent entrepreneurs were
initially identified from participants who were registered as monitors of Yahoo Japan Value
Insight in two phases. In the first phase, a question in the first screening was designed to
identify people who wish to start businesses.

In the future, do you want to start your own business?
A respondent could choose from four options to the question (yes to great extent; yes,
3
The databases taking the approach of screening from a large number of candidates of nascent entrepreneurs
include the Panel Study of Entrepreneurial Dynamics (PSED). The PSED employed a market research company
to screen a total of 64,622 households in the United States using the survey procedure described in detail in
Schjoedt and Shaver (2007). Despite the relatively large size of population, the PSED ended up collecting 553
responses from nascent entrepreneurs.
9
Business startup reasons
but only if it is possible to do so; no extent; I have already started my business). Only
individuals falling into “yes to great extent” and “yes, but only if it is possible to do so” were
considered for the second screening.
The following question served as the second screening, designed to identify people
who were prepared to start businesses in very near future.

Have you already created your business plan, or developed your business idea?

Are you planning to start your business within three years?
A respondent could answer “no” or “yes” to the questions. The affirmative answers
to both questions were necessary for respondents to be included in the study. In the second
phase of the research, 3,247 respondents who satisfied the inclusion criteria described above
were sent a detailed web questionnaire with a response rate of 74.6 percent.
Sample
From effective responses, I constructed a sample for current analysis, guided by two
principles. First, because the behaviors and motives of serial entrepreneurs are likely to be
different from the first-time entrepreneurs, I restricted the analysis to the first transition to
business ownership. Second, to avoid self-justification bias and survival bias, employees
spending a long time after they had the first intension to start a business should not be used
for analysis. As a result, the sample consists of employees who (a) have no prior experience
of business ownership; (b) have spent less than five years ever since they had intension to
start a business; (c) are employed regularly on a full-time basis, including middle managers
and general employees; (d) do not work for the public sector and primary industry
(agriculture and extracting industries).
The timeframe of five years was used because it could cover the almost all possible
range of the time when nascent entrepreneurs spend to put their businesses on the track. The
10
Business startup reasons
average time necessary for gestation period has been found to be one year, and 90%
completed the gestation within three years (Reynolds and Miller, 1992).
I examined a demographic distribution of the sample of employees planning to start
businesses used for this study and general statistics of regular employees available from
Labor Force Survey, a nationally representative sample of work force in Japan. Comparing
structures of age and gender indicates that the individuals in their 30s and 40s are
overrepresented in my sample of employees who plan to start businesses. This might be
reflecting on overrepresentation of people who plan to start businesses in these age groups
and also younger generations, who are likely to be registered as monitors of web-based
research.
Measures of Business Startup Reasons
Evaluations of respondents for ten items of startup reasons were subjected to a principal
component analysis to identify latent needs for starting a business. Three items represent a
lack of esteem from the employer (dismal prospect for organization, frustration with wage,
frustration with career prospect). Two items represent a need for self fulfillment (to make the
best use of technological specialty and knowledge, to develop an innovative idea). Two items
represent a need for autonomy in work (free to adopt my approach to work, get a greater
flexibility for personal life). Two items represent necessity to earn a livelihood (to be
respected in society, to earn livelihood income). Finally, I also added an item (to earn a larger
income) related to a desire for financial success.
All items were asked by the question, “to what extent are the following reasons
important to you in starting your own business?” Responses were given on a one-to-five
scale: one, “to no extent”; two, “little extent”; three, “some extent”; four, “great extent”; five,
“to a very great extent”.
11
Business startup reasons
Scores for rating to these reasons items were subjected to a principal component
factor analysis. The results of the analysis will be reported in next section. The produced
scores for motive factors will be then used as dependent variables of the model of hypothesis
test.
Measures of Work Conditions
Firm size. I measured firm size as the number of workers in the firm. Because the questions
which appear in the survey asked respondents to indicate which of 5 size-bands describes
employer, I used dummy variables to differentiate between different size-categories. Five size
categories were used: less than 20 workers, 20 to 99 workers, 100 to 299 workers, 300 to 999
workers, and 1000 workers or more.
Firm age. The age of the firm is measured as years elapsed since the establishment of the
firm. Because the question which appears in the survey asked respondents to indicate which
of three age-bands represented employer, dummy variables were used. I used three age
categories: 27 years or younger, 28 to 62 years, 63 years or older.
Tenure. I measured tenure of employees as the years of current job experience. Because
respondents were asked to select which of 9 bands represented job tenure, I used dummy
variables. Because a distribution of job tenure in the sample is skewed toward employees
with tenures of less than 20 years, I settled on four tenure categories: 5 years or less, 6 to 10
years, 11 to 15 years, and 20 years or more.
Control Variables
To complete the model of startup motive, I included a set of control variables. The model
included a dummy variable coded one for females (female), the number of years of education
completed (years of education), dummy variables for the person’s age at the time of survey
(individual’s age: 20-29 years old; individual’s age: 30-39 years old; individual’s age: 40-49
12
Business startup reasons
years old), and a dummy variable coded one for married persons (married).
The model also included the amount of wage and salary income received in a year
(yearly wage). Because the questions which appear in the survey asked respondents to select
which of four bands describes the amount of income, dummy variables were used. I used four
categories: 2.5 million yen or less, 2.5 million to 5 million yen, 5 million to 10 million yen,
10 million yen or more.
It is also important to separate the effect of duration of current job and the effect of
mobility of a person in her entire career history. The model thus included a measure of career
mobility by counting the number of job changes since leaving school (number of job
changes). The set of control variables also included a dummy variable coded one for middle
managers to account for difference in job responsibility (middle manager).
Estimation Strategy
I used a multivariate regression method to test the effects of firm size on latent motives.
Particular personality of persons which is unobserved to researchers, such as a preference for
autonomy in workplace that affects one startup motive has an effect on the other. For
example, preference of a person for autonomy that makes her consider flexibility in work
schedule as more important may be connected closely with an innate basis for self fulfillment.
Psychological studies have documented a number of common traits among successful
entrepreneurs (for a review, see Begley and Boyd, 1987). Accounting for personality traits in
the models with multiple dependent variables is always challenging, but methodologically,
the most widespread method of addressing this sort of problem is to use a multivariate
regression method, which analyzes multiple dependent variables jointly in a single statistical
method rather than each separately.
13
Business startup reasons
Descriptive Statistics
The descriptive statistics in Table 1 presents an overview of the sample used for analysis.
One in five persons was female, and the average person received 18 years of education. The
most frequent category for individual’s age was age of 30s, reflecting that the sample is
skewed toward younger generations, and six in ten were married. The most frequently
occurring group of tenure was employees with tenures of five years or less, and the most
frequent category of firm size was firms with less than 20 workers. The most frequent
category of the age of firms was 27 years or less. 35 percent of employees held middle
management positions, and the average number of job changes a person had was 1.7. The
most frequent category of the amount of wage or salary income in a year was from 2.5
million to 5 million yen.
RESULTS
I first subjected ten reason items to a principal components analysis (varimax rotation), where
the eigenvalue was greater than one. The analysis produced four factors, accounting for a
total of 70.1 percent of the variance. As shown in Table 2, the first factor, a need to solve
career problems, involved four items, the second, a need for work flexibility involved two
items, the third, a need for self fulfillment involved two items, and a need for livelihood
security had two items4. To predict dependent variables, I calculated scores for each of the
four reason scales.
The predicted factor scores showed consistency with measures of overall life and job
satisfaction scores available from survey data. As shown in Table 3, the correlation
coefficients of job satisfaction score with a need to solve career problem, a need for work
flexibility, and a need for livelihood security were -0.29, -0.20, -0.15 respectively, and all of
4
The first factor, career problem involves the item, “earn a larger income”. However, loading coefficient is
0.34, which is significantly lower than the coefficients of other items in the same factor.
14
Business startup reasons
coefficients were significant statistically, whereas the correlation coefficient between job
satisfaction score and a need for self fulfillment was not significant statistically. This suggests
that a need to solve career problem, a need for work flexibility, and a need for livelihood
security were associated with a low level of job satisfaction. The correlation coefficients of
the life satisfaction score with a need to solve career problems, a need to secure livelihood
were -0.15 and -0.10 respectively, and both were significant. The need for work flexibility
and that for self fulfillment were not significantly correlated with the life satisfaction score.
This suggests that a need to solve career problem and that to secure livelihood is associated
with inadequacy of satisfaction with job and life.
Table 4 presents model estimates of major startup motives using a multivariate
regression method. Model 1 included all variables except for firm-size and firm-age
covariates. The effect of individuals in their 20s and 30s on a need to secure livelihood were
negative and significant. This implies that for employees who wish to start businesses, those
of 50 years old or more have a larger need to secure livelihood as compared to younger
generations. The effect of married persons on a need for self fulfillment was significant and
positive, implying that people planning to start businesses have a large need for self
fulfillment if they are married. The effects of wage income of 2.5 million yen or less and that
of 2.5 million to 5 million yen on a need to solve career problem were positive and significant,
implying that career problems are often expressed by low-income earners. The effect of
employees with job tenure of five years or less, and those with job tenure of six to ten years
on a need for self fulfillment were positive and significant, implying that short-tenured
employees are more likely to stress that kind of need as a startup motive.
I tested proposed hypotheses using Model 2 that includes all covariates, and this
increases model fit considerably. Firms having less than 20 workers, those having 20-99
workers, those having 100-299 workers, and those having 300-999 workers have positive and
15
Business startup reasons
significant effects for a need to solve career problem. This implies a need to solve career
problem is higher among employees of small firms planning to start businesses (lending
support to Hypothesis 1b). On the other hand, the effects of small firms were significant
neither for a need for self fulfillment (not offering support for Hypothesis 1a), for a need for
work flexibility (not offering support for Hypothesis 1c) nor for a need to secure livelihood
(not offering support for Hypothesis 1d). The joint effects of firms having less than 20
workers, firms having 20-99 workers, firms having 100-299 workers, and firms having
300-999 workers on a need to solve career problem were significant (F = 2.25; p < 0.10).
This suggests that employees of small firms who wish to start businesses have higher level of
needs to solve career problems as compared to those of large firms. On the other hand, there
was no significant difference between employees of large and small firms in both the level of
a need for self fulfillment and the level of a need for work flexibility. The effects of young
firms were statistically significant for neither of startup motives.
Table 5 presents separate reestimates of this model for subsamples of middle
managers and general employees. In Model 1, I used a subsample of middle managers to test
the effects of firm size, while in Model 2 I used a subsample of general employees. In Model
1, the effects of firms with less than 20 workers and of firms with 20 to 99 workers on a need
to solve career problem were positive and significant, whereas in model 2, the effect of firm
size on a need to solve career problem is limited for those working for firms with less than 20
workers. This suggests that the effects of small firms on a need to solve career problem is
associated with the tendency for middle managers.
DISCUSSION
The results of this study provide a strong support for the explanation based on the notion that
the reason for why small firms produce entrepreneurs is that employees of small firms are
16
Business startup reasons
frustrated with their career prospects. The results from multivariate regression models
suggested that those working for small firms are more likely to stress a need to solve career
problems as a major startup motive. This conclusion is particularly robust when it comes to a
motive of middle managers of small firms. This may reflects that a path to levels above
middle management becomes less clear in small firms. The finding can also be interpreted as
the consequence of exclusion of employees of small firms from labor market.
The results have not offered compelling support to the self-selection perspective.
This is not, however, the same as meaning that this study was effective in accounting for all
possible patterns of self-selection behavior. The preference of individuals explored in this
study was based on a desire for autonomy. However, there has been an alternative view that
less risk-averse individuals are more willing to become employees of small firms on the
assumption that small firms on average pay more variable wages (Parker, 2009). Future
research need to explore this issue empirically. It could also examine whether workers who
left small firms create more or less successful ventures compared to those who left large
firms.
I would not consider it right to generalize about characteristics of entrepreneurs from
the results of this paper. Clearly, the sample population from which data were drawn was not
generally representative. First, depth of analysis gained through the measurement of detailed
startup reasons, and data of job and workplace characteristics comes at the expense of breadth
that could be achieved through a larger sample of employers and employees. One might
worry, for example, the generalizablity of results toward older generations underrepresented
in the sample of this study. Second, approach to focus on general workforce regardless of
industry limited the ability of this study to examine how market and technology
characteristics might interact with the size of firms and major motives of employees to start
businesses. For example, a more detailed analysis of emerging and declining industries might
17
Business startup reasons
demonstrate the significance of the industry structure, the stage of industry life cycle, and
technological regime.
Managerial implications from the findings of this study is that small employers who
seek to retain employees need to be aware that potential quitters who wish to setting up their
own businesses could be better found among middle managers frustrated with their career
opportunities. The findings of the study also have several implications for policy makers
seeking to promote entrepreneurship. The empirical regularity between small firms and the
rate of employees leaving to start their own businesses has suggested potential benefits of
considering the indirect effects of policies designed to support and sustain existing small
firms (Sørensen, 2007). For example, such policies have been considered desirable because
they may not only promote the development of existing small firms but also help increase the
supply of entrepreneurial talent in an economy. Based on this notion of small firms inspiring
persons to pursue entrepreneurial opportunities, educators and activists who seek to develop
an entrepreneurial mind-set among students have introduced programs designed to support
entrepreneurship internship for employees in small companies. However, the findings of this
study suggested that employees of Japanese small firms start a business mostly because they
are frustrated with their careers opportunities, and the findings offered little compelling
evidence of the transmission theory. This suggests that the potential effectiveness of such
kind of programs can be compromised unless the selection of firms for internship programs is
managed correctly.
REFERENCES
Acs, Z. J., Catherine, A. & Robb, A. (1999). Measures of job flow dynamics in the U.S. (U.S.
Bureau of the Census Center for Economic Studies Discussion Paper 99-1).
Audretsch, D. B. & Acs, Z. J. (1990). The entrepreneurial regime, learning, and industry
18
Business startup reasons
turbulence, Small Business Economics, 2, 119-128
Bates, T. (1990). Entrepreneur human capital inputs and small business longevity. Review of
Economics and Statistics, 72, 551-559.
Begley, T. M. & Boyd, D. P. (1987). Psychological characteristics associated with
performance in entrepreneurial firms and smaller businesses. Journal of Business
Venturing, 2, 79-93.
Birch, D. L. (1987). Job Creation in America. Free Press, New York.
Busenitz, L. W. & Barney, J. B. (1997). Differences between entrepreneurs and managers in
large organizations: Biases and heuristics in strategic decision-making. Journal of
Business Venturing, 12, 9-30.
Carter, N. M., Gartner, W. B., Shaver, K. G. & Gatewood, E. J. (2003). The career reasons of
nascent entrepreneurs, Journal of Business Venturing, 18, 13-39.
Choi, Y.R., & Phan, P.H. (2006). The influences of economic and technology policy on the
dynamics of new firm formation, Small Business Economics, 26, 493-503.
Dobrev, S. D. & Barnett, W. P. (2005). Organizational roles and transition to
entrepreneurship. Academy of Management Journal, 48, 433-449.
Fritsch, M. (1992). Regional differences in new firm formation: evidence from West
Germany. Regional Studies, 26, 233-241.
Gerlach, K., & Wagner, J. (1994). Regional differences in small firm entry in manufacturing
industries: Lower Saxony, 1979-1991. Entrepreneurship and Regional Development,
6, 63-80.
Gompers, P., Lerner, J. & Scharfstein, D. S. (2005). Entrepreneurial spawning: public
corporations and the genesis of new ventures, 1986 to 1999. Journal of Finance, 60,
577-614.
Harada, N. (2003). Who succeeds as entrepreneur? an analysis of the post-entry performance
19
Business startup reasons
of new firms in Japan. Japan and the World Economy, 15, 211-222.
Honjo, Y. (2004). Growth of new start-up firms: evidence from the Japanese manufacturing
industry. Applied Economics Letters, 11, 21-32.
Jovanovic, B. (1979). Job matching and the theory of turnover. Journal of Political Economy,
87, 972-990.
Klepper, S. (2002).The capabilities of new firms and the evolution of the US automobile
industry. Industrial and Corporate Change, 11, 645-666.
Lazear, E. P. (2005). Entrepreneurship. Journal of Labor Economics, 23, 649-680.
Newbert, S., Kirchhoff, B.A. & Walsh, S.T. (2007). Defining the relationship among
founding resources, strategies, and performance in technology-intensive new
ventures: evidence from the semiconductor silicon industry. Journal of Small
Business Management, 45, 438-466.
Okamuro, H. (2008). How different are the regional factors of high-tech and low-tech
start-ups? evidence from Japanese manufacturing industries. International
Entrepreneurship and Management Journal, 4, 199-215.
Parker, S. C. (2009). Why do small firms produce the entrepreneurs? Journal of
Socio-Economics, 38, 484-494.
Rees, H. & Shah, A. (1986). An empirical analysis of self-employment in the U.K. Journal of
Applied Econometrics, 1, 95-108.
Reynolds, P. & Miller, B. (1992). New firm gestation: conception, birth, and implications for
research. Journal of Business Venturing, 7, 405-417.
Reynolds, P. D., Storey, D. J. & Westhead, P. (1994). Cross-national comparisons of the
variation in new firm formation rates. Regional Studies, 28, 443-456.
Schjoedt, L. & Shaver, K. G. (2007). Deciding on an entrepreneurial career: a test of the pull
and push hypotheses using the Panel Study of Entrepreneurial Dynamics data.
20
Business startup reasons
Entrepreneurship Theory and Practice, 31, 733-752.
Small and Medium Enterprise Agency (2006; 2009). White Paper on Small and Medium
Enterprise in Japan. Japan Small Business Research Institute: Tokyo, Japan.
Sonobe, T., Kawakami, M. & Otsuka, K. (2003). Changing roles of innovation and imitation
in industrial development: the case of the machine tool industry in Taiwan.
Economic Development and Cultural Change, 52, 103-128.
Sørensen, J. B. (2007). Bureaucracy and entrepreneurship: workplace effects on
entrepreneurial entry. Administrative Science Quarterly, 52, 387-412.
Spilling, O. R. (1996). Regional variation of new firm formation: the Norwegian case.
Entrepreneurship and Regional Development, 8, 217-243.
Troske,
K.
R.,
(1999).
Evidence
on
the
employer
size-wage
premium
from
worker-establishment matched data. Review of Economics and Statistics, 81, 15-26.
Tushman, M. L. & Anderson, P. (1986). Technological discontinuities and organizational
environments. Administrative Science Quarterly, 31, 439-465.
Tsuchiya, R. (2009). Why Do Small Firms Produce More Entrepreneurs? A Test of the Pull
and Push Hypotheses. Paper presented at the International Council for Small
Business 2009 World Conference, Seoul, Korea, June 23.
Van Stel, A., Storey, D. J. & Thurik, A. R. (2007). The effect of business regulations on
nascent and young business entrepreneurship, Small Business Economics, 28,
171-186.
Yahata, Sigemi (1998). Former employees as owners of small independent businesses,
Monthly Journal of the Japan Institute of Labor, 40, 2-14 (in Japanese).
Yasuda, T. (2005). Firm growth, size, age and behavior in Japanese manufacturing. Small
Business Economics, 24, 1-15.
21
Business startup reasons
Table 1
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
22
Descriptive statistics and correlations†
Female
Years of education
Individual's age: 20-29 years old
Individual's age: 30-39 years old
Individual's age: 40-49 years old
Married
Tenure: 5 years or shorter
Tenure: 6-10 years
Tenure: 11-15 years
Middle manager
Number of job changes
Yearly wage: 2.5M Yen or less
Yearly wage: 2.5M-5M Yen
Yearly wage: 50M-100M Yen
Firm size: 1-19 workers
Firm size: 20-99 workers
Firm size: 100-299 workers
Firm size: 300-999 workers
Firm age: 27 years or younger
Firm age: 28-62 years
Mean
0.18
17.55
0.11
0.46
0.33
0.59
0.38
0.24
0.19
0.35
1.65
0.10
0.45
0.39
0.33
0.23
0.14
0.13
0.44
0.38
S.D.
0.39
2.05
0.31
0.50
0.47
0.49
0.48
0.43
0.39
0.48
1.73
0.30
0.50
0.49
0.47
0.42
0.35
0.34
0.50
0.49
Min
0.00
12.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
Max
1.00
21.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
10.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.
0.02
0.04
0.05
-0.07
-0.15
0.12
-0.06
0.01
-0.21
0.06
0.13
0.09
-0.16
0.00
-0.07
0.07
0.07
0.04
-0.08
2.
-0.02
0.05
-0.06
0.00
0.06
0.02
-0.02
0.00
-0.25
-0.08
-0.16
0.12
-0.21
-0.04
-0.03
0.14
-0.09
0.03
3.
-0.33
-0.25
-0.29
0.40
-0.13
-0.17
-0.14
-0.12
0.16
0.10
-0.16
0.00
0.00
0.03
-0.03
0.06
-0.03
4.
-0.65
-0.10
-0.02
0.27
0.06
-0.21
-0.02
0.01
0.12
-0.07
0.03
0.01
0.00
0.05
0.06
-0.02
5.
0.19
-0.18
-0.12
0.06
0.18
0.08
-0.15
-0.11
0.16
-0.02
-0.03
-0.03
-0.04
-0.07
0.03
6.
-0.26
0.01
0.05
0.25
-0.03
-0.13
-0.16
0.19
-0.08
0.01
0.00
-0.04
-0.13
0.04
7.
-0.44
-0.37
-0.20
0.29
0.09
0.17
-0.17
0.08
0.08
-0.03
-0.05
0.30
-0.19
8.
-0.27
-0.05
0.01
-0.01
0.08
-0.09
0.02
-0.01
0.02
-0.03
-0.04
0.06
9.
0.12
-0.09
-0.04
-0.06
0.12
-0.01
-0.05
-0.02
0.06
-0.06
0.03
10.
-0.01
-0.21
-0.22
0.27
-0.04
-0.05
0.06
-0.03
-0.09
0.06
Business startup reasons
Table 1 (continued)
11.
12.
13.
-0.03
0.13
-0.30
-0.09
-0.26
-0.73
0.21
0.14
0.10
0.02
0.00
0.10
-0.05
-0.04
0.04
-0.14
-0.04
-0.05
0.23
-0.04
0.12
-0.12
0.05
-0.05
†
S.D.: standard deviations
23
14.
-0.10
-0.07
-0.04
0.08
-0.04
0.01
15.
-0.39
-0.29
-0.28
0.22
-0.06
16.
-0.22
-0.22
0.03
0.02
17.
-0.16
0.03
0.03
18.
-0.11
0.02
19.
-0.70
Business startup reasons
Table 2
Factor loadings for reasons items: n = 294†
Factor
Variance
1
2
3
4
Career
Work
Self
Livelihood
problem
flexibility
fulfillment
Security
2.43
1.665
1.554
1.361
Percentage variance accounted for
0.243
0.167
0.155
0.136
To earn a larger income
0.338
Dismal prospect for organization
0.524
Frustration with wage
0.547
Frustration with career prospect
0.481
Free to adapt my approach to work
-0.341
0.479
To get a greater flexibility for personal
life
0.739
To make the best use of technological
specialty and knowledge
0.624
To develop an innovative idea
0.701
To earn livelihood income
To be respected in society
†
Factor loadings smaller than 0.30 have been suppressed.
24
0.720
0.347
0.532
Business startup reasons
Table 3 Correlations of factor scores with overall job and life
satisfaction scores
Career problem
Work flexibility
Self fulfillment
Livelihood
Job satisfaction
security
score
Career problem
*
Self fulfillment
0.42
***
Work flexibility
0.28
***
0.31
***
Livelihood security
0.20
***
0.14
**
0.08
Job satisfaction score
-0.29
***
-0.20
***
0.00
-0.15
**
Life satisfaction score
-0.15
***
-0.09
0.09
-0.10
*
p < .10; ** p < .05, *** p < .01; two-tailed tests.
25
0.43
***
Business startup reasons
Table 4 Multivariate multiple regression models of major motives for starting new
businesses†
Model 1
Female
Years of education
Model 2
Career
Work
Self
Livelihood
Career
Work
Self
Livelihood
problem
flexibility
fulfillment
security
problem
flexibility
fulfillment
security
-0.396
0.216
0.123
-0.115
-0.373
0.185
0.113
-0.088
(0.261)
(0.216)
(0.204)
(0.187)
(0.261)
(0.217)
(0.207)
(0.189)
0.070
-0.019
0.035
0.010
0.107**
-0.023
0.041
0.020
(0.053)
(0.044)
(0.041)
(0.038)
(0.054)
(0.045)
(0.043)
(0.039)
-0.092
0.021
0.473
-0.755**
0.222
-0.079
0.495
-0.685*
(0.498)
(0.412)
(0.391)
(0.358)
(0.508)
(0.422)
(0.403)
(0.368)
-0.050
-0.003
-0.045
-0.658**
0.091
-0.052
-0.028
-0.608**
(0.389)
(0.322)
(0.305)
(0.279)
(0.391)
(0.325)
(0.311)
(0.284)
0.202
0.321
0.033
-0.291
0.399
0.270
0.063
-0.241
(0.376)
(0.311)
(0.295)
(0.270)
(0.380)
(0.316)
(0.302)
(0.276)
0.211
0.176
0.603***
-0.135
0.280
0.173
0.623***
-0.125
(0.214)
(0.177)
(0.167)
(0.153)
(0.214)
(0.178)
(0.170)
(0.155)
0.956*
0.164
0.079
-0.141
0.619
0.308
0.120
-0.138
(0.556)
(0.460)
(0.436)
(0.399)
(0.570)
(0.474)
(0.453)
(0.413)
0.900**
-0.123
-0.057
0.059
0.591
-0.022
-0.057
0.013
(0.449)
(0.372)
(0.352)
(0.323)
(0.460)
(0.382)
(0.365)
(0.333)
0.474
-0.173
0.134
-0.525*
0.240
-0.045
0.153
-0.542*
(0.425)
(0.352)
(0.333)
(0.305)
(0.431)
(0.359)
(0.343)
(0.313)
-0.288
0.381
0.500*
-0.145
-0.443
0.481
0.498*
-0.198
(0.371)
(0.307)
(0.291)
(0.266)
(0.377)
(0.314)
(0.300)
(0.274)
-0.090
0.395
0.640**
0.136
-0.196
0.435
0.627**
0.107
(0.348)
(0.288)
(0.273)
(0.250)
(0.349)
(0.290)
(0.277)
(0.253)
-0.323
-0.254
0.260
0.284
-0.343
-0.230
0.256
0.272
(0.334)
(0.276)
(0.262)
(0.240)
(0.332)
(0.276)
(0.264)
(0.241)
-0.195
0.084
-0.056
-0.321**
-0.199
0.051
-0.071
-0.326**
(0.218)
(0.180)
(0.170)
(0.156)
(0.217)
(0.181)
(0.173)
(0.158)
-0.001
-0.044
0.035
-0.077
0.000
-0.026
0.043
-0.073
(0.070)
(0.058)
(0.055)
(0.050)
(0.070)
(0.058)
(0.056)
(0.051)
0.971***
-0.174
0.046
0.033
(0.337)
(0.280)
(0.268)
(0.245)
0.775**
-0.079
0.073
0.235
(0.337)
(0.280)
(0.268)
(0.244)
0.842**
0.374
0.374
0.254
Individual's age: 20-29 years
old
Individual's age: 30-39 years
old
Individual's age: 40-49 years
old
Married
Yearly wage: 2.5M Yen or
less
Yearly wage: 2.5M-5M Yen
Yearly wage: 50M-100M Yen
Tenure: 5 years or shorter
Tenure: 6-10 years
Tenure: 11-15 years
Middle manager
Number of job changes
Firm size: 1-19 workers
Firm size: 20-99 workers
Firm size: 100-299 workers
26
Business startup reasons
Firm size: 300-999 workers
Firm age: 27 years or younger
Firm age: 28-62 years
Constant
(0.350)
(0.291)
(0.278)
(0.254)
0.580*
-0.123
0.055
-0.159
(0.350)
(0.291)
(0.278)
(0.254)
0.047
-0.294
-0.031
0.065
(0.295)
(0.245)
(0.235)
(0.214)
-0.041
-0.098
-0.024
0.015
(0.269)
(0.223)
(0.213)
(0.195)
-1.670
0.100
-1.669*
0.999
-2.896**
0.254
-1.884*
0.671
(1.214)
(1.004)
(0.952)
(0.872)
(1.281)
(1.065)
(1.018)
(0.929)
R2
0.07
0.07
0.11
0.14
0.11
0.10
0.12
0.16
F
0.94
0.96
1.47
2.09
1.11
1.02
1.24
1.80
Significance of F
0.55
0.51
0.08
0.00
0.32
0.44
0.20
0.01
Breusch-Pagan test of
independence: χ2
129.00
131.83
Significance of χ
0.00
0.00
294
294
2
Number of observations
*
p < .10;
†
**
p < .05,
***
p < .01; two-tailed tests.
Standard errors are in parentheses. All models include dummy variables for industry. .
27
Business startup reasons
Table 5 Multivariate multiple regression models of major motives for starting new
businesses: middle managers and general employees†
Female
Years of education
Model 1
Model 2
Middle managers only
General employees only
Career
Work
Self
Livelihood
Career
Work
Self
Livelihood
problem
flexibility
fulfillment
security
problem
flexibility
fulfillment
security
-0.676
0.309
-0.597
-0.349
-0.244
0.242
0.230
-0.094
(0.794)
(0.700)
(0.640)
(0.559)
(0.291)
(0.225)
(0.227)
(0.207)
0.065
-0.055
0.040
-0.062
0.122*
-0.044
0.070
0.070
(0.095)
(0.084)
(0.077)
(0.067)
(0.072)
(0.056)
(0.056)
(0.052)
0.596
-0.107
0.854
-0.646
0.324
0.405
0.464
-0.573
(1.112)
(0.981)
(0.896)
(0.783)
(0.709)
(0.549)
(0.553)
(0.505)
-0.431
-0.611
0.131
-0.397
0.615
0.543
0.001
-0.620
(0.664)
(0.585)
(0.535)
(0.467)
(0.587)
(0.454)
(0.458)
(0.418)
-0.131
-0.426
-0.121
-0.010
0.968
0.915**
0.183
-0.290
(0.569)
(0.502)
(0.459)
(0.401)
(0.599)
(0.464)
(0.467)
(0.427)
0.387
0.622*
0.690**
-0.330
0.242
-0.008
0.569***
-0.082
(0.411)
(0.363)
(0.331)
(0.290)
(0.266)
(0.205)
(0.207)
(0.189)
1.003
-2.294
0.965
-1.624
0.851
0.330
0.316
-0.578
(2.054)
(1.811)
(1.654)
(1.446)
(0.780)
(0.604)
(0.608)
(0.556)
0.385
-0.584
-0.370
-0.090
0.852
-0.002
0.294
-0.323
(0.750)
(0.662)
(0.604)
(0.528)
(0.696)
(0.538)
(0.543)
(0.496)
0.206
0.226
0.208
-0.394
0.351
-0.472
0.371
-1.059**
(0.646)
(0.570)
(0.520)
(0.455)
(0.694)
(0.537)
(0.541)
(0.495)
-0.760
0.760
0.536
-0.524
-0.307
0.386
0.398
-0.076
(0.652)
(0.575)
(0.525)
(0.459)
(0.501)
(0.388)
(0.391)
(0.357)
-0.681
0.665
0.945*
-0.319
-0.034
0.514
0.379
0.196
(0.671)
(0.592)
(0.541)
(0.473)
(0.455)
(0.352)
(0.355)
(0.324)
-0.379
0.064
0.242
0.111
-0.459
-0.415
0.043
0.300
(0.539)
(0.476)
(0.435)
(0.380)
(0.471)
(0.364)
(0.367)
(0.335)
0.130
-0.066
0.203*
0.005
-0.077
-0.049
-0.043
-0.120*
(0.139)
(0.122)
(0.112)
(0.098)
(0.088)
(0.068)
(0.069)
(0.063)
1.249**
0.187
-0.128
0.149
0.816*
-0.168
0.349
0.399
(0.630)
(0.556)
(0.508)
(0.444)
(0.461)
(0.357)
(0.359)
(0.328)
1.301**
0.281
-0.304
0.173
0.497
-0.089
0.412
0.554*
(0.625)
(0.551)
(0.504)
(0.440)
(0.439)
(0.340)
(0.342)
(0.313)
0.867
0.263
0.236
0.413
0.760
0.458
0.553
0.501
(0.599)
(0.528)
(0.483)
(0.422)
(0.483)
(0.374)
(0.377)
(0.344)
0.929
0.256
0.201
0.431
0.364
-0.217
0.015
-0.214
Individual's age: 20-29 years
old
Individual's age: 30-39 years
old
Individual's age: 40-49 years
old
Married
Yearly wage: 2.5M Yen or less
Yearly wage: 2.5M-5M Yen
Yearly wage: 50M-100M Yen
Tenure: 5 years or shorter
Tenure: 6-10 years
Tenure: 11-15 years
Number of job changes
Firm size: 1-19 workers
Firm size: 20-99 workers
Firm size: 100-299 workers
Firm size: 300-999 workers
28
Business startup reasons
Firm age: 27 years or younger
Firm age: 28-62 years
Constant
(0.682)
(0.602)
(0.550)
(0.480)
(0.446)
(0.345)
(0.348)
(0.318)
-0.149
-0.824*
-0.036
0.380
0.154
-0.036
-0.153
-0.219
(0.547)
(0.482)
(0.441)
(0.385)
(0.402)
(0.311)
(0.314)
(0.286)
-0.306
-0.609
-0.107
0.040
0.139
0.187
-0.142
-0.122
(0.484)
(0.426)
(0.390)
(0.340)
(0.356)
(0.275)
(0.277)
(0.253)
-1.591
1.476
-1.862
1.665
-3.946**
0.010
-2.766**
0.097
(2.240)
(1.975)
(1.805)
(1.577)
(1.734)
(1.342)
(1.352)
(1.235)
R2
0.22
0.21
0.29
0.26
0.14
0.21
0.13
0.24
F
0.78
0.70
1.12
0.94
1.01
1.60
0.95
1.98
Significance of F
0.77
0.85
0.34
0.56
0.46
0.04
0.54
0.01
Breusch-Pagan test of
independence: χ2
60.56
84.44
Significance of χ2
0.00
0.00
101
193
Number of observations
*
p < .10;
†
**
p < .05,
***
p < .01; two-tailed tests.
Standard errors are in parentheses. All models include dummy variables for industry.
29